For many years, heat shock or stress proteins have been regarded as intracellular molecules that have a range of housekeeping and cytoprotective functions, only being released into the extracellular environment in pathological situations such as necrotic cell death. However, evidence is now accumulating to indicate that, under certain circumstances, these proteins can be released from cells in the absence of cellular necrosis, and that extracellular heat shock proteins have a range of immunoregulatory activities. The capacity of heat shock proteins to induce pro-inflammatory responses, together with the phylogenetic similarity between prokaryotic and eukaryotic heat shock proteins, has led to the proposition that these proteins provide a link between infection and autoimmune disease. Indeed, both elevated levels of antibodies to heat shock proteins and an enhanced immune reactivity to heat shock proteins have been noted in a variety of pathogenic disease states. However, further evaluation of heat shock protein reactivity in autoimmune disease and after transplantation has shown that, rather than promoting disease, reactivity to self-heat shock proteins can downregulate the disease process. It might be that self-reactivity to heat shock proteins is a physiological response that regulates the development and progression of pro-inflammatory immunity to these ubiquitously expressed molecules. The evolving evidence that heat shock proteins are present in the extracellular environment, that reactivity to heat shock proteins does not necessarily reflect adverse, pro-inflammatory responses and that the promotion of reactivity to self-heat shock proteins can downregulate pathogenic processes all suggest a potential role for heat shock proteins as therapeutic agents, rather than as therapeutic targets.